Study Of Road Detection From Remote Sensing Image Based On Mathematical Morphology

With the development of space technology, sensor technology, computer science and technology etc, information collected through remote sensing has been one of the main sources for geographical data. Compare to the traditional ways of geographical data collection, remote sense has great advantage at many aspects. How to make best use of remote sensing data has been widely discussed in the field of global information science around the world. Road network information is one of the most important parts of geographical information. Due to its importance in regional planning, traffic navigation and military spying, road network detection from remote sensing image has drawn great attention among researchers all over the world, and plenty of achievements have been made. With the development of airborne remote sense and space remote sense, it comes to reality to get abundant of data with high space and temporal resolution. However, due to the limitation of artificial intelligence and complexity of the remote sensing data, there is still no universal way of automatic road detection from remote sensing image. And the most popular way of getting road information is through semi-automatic method based on interaction between people and computer. To a large extent, it increases the cost and reduces the productivity.Mathematical morphology (MM) belongs to application mathematics. It has been developed over 30 years since being discovered in Paris. As a theoretical model for digital images built upon lattice theory, Mathematical morphology considers digital images as geometrical objects, to be analyzed through their interactions with other geometrical objects. It can analysis and manipulate digital image quantitively and has produced fast and efficient algorithms for computer analysis of images, meanwhile, It has found applications in bio-medical imaging, materials science, geosciences, remote sensing, quality control and data analysis etc.The thesis contains five chapters: Chapter one discusses the theoretical and realistic meaning of the study, and outlines the history of research on road detection from remote sensing image; chapter two studies the basic theory of Mathematical morphology; Chapter three discusses the fundamental problem in the road network detection, and analyses the availability and limitation of application of Mathematical morphology in road network detection from remote sensing image; Chapter four tests the method mentioned above on the real remote sensing data; Chapter five gives conclusion and discussion.